Hi there,

I have am getting different outputs after running enrichGO on cluster profiler when I put the same genes into enrichR (by Maayan Lab) website. Example here using Biological Process 2021

I have also tried other GO terms (Cellular Component and Molecular Funtion giving the same discordant results)

Code should be placed in three backticks as shown below

input_marker_files <- 'input_file'
for(i in 1:1){
  markerFile <- paste0(input_marker_files, i,'.tsv')
  marker.gene.clust[[i]] <- read.table(markerFile)
  marker.gene.clust.filt <- marker.gene.clust[[i]][marker.gene.clust[[i]]$summary.logFC>1.5,]
  #print(markerFile)


  gene <- marker.gene.clust.filt$Symbol

  gene.df <- bitr(gene, fromType = "SYMBOL",
        toType = c("ENSEMBL", "ENTREZID"),
        OrgDb = org.Hs.eg.db)

ego <- enrichGO(gene         = gene.df$ENSEMBL,
                OrgDb         = org.Hs.eg.db,
                keyType       = 'ENSEMBL',
                ont           = "BP",
                pAdjustMethod = "BH",
                pvalueCutoff  = 0.01,
                qvalueCutoff  = 0.05)



p2  <-  barplot(ego, showCategory=10) +
  theme(axis.text.y= element_text(size=12)) +
  scale_y_discrete(guide = guide_axis(n.dodge=1)) +
   ggtitle(paste0("Epithelial Cluster ", i))

p2

Thanks

What I have noticed, while using an Appyter for enrichR, is that even between the local result I get on the notebook is different from the one that I would get if I input the same list directly on the EnrichR site. I think it is because on the site it is using the most update GO list from 2021, while on my notebook its the 2018 one. Also, my result are not that different, just arranged in a different way, but mostly the same terms. I will test if this changes after I select the same version for both.

As for your question, I wonder if something like this could be happening to you. It would be worth digging into clusterprofiler github or any other source to know what lists of GO terms it is using for the version you are using.